Drug resistance is a significant factor in tumor chemotherapy and is also a useful marker for the genetic analysis of mammalian cells. Resistance to adenine analogs in somatic cells in vitro is most frequently associated with a lack of or alterations in adenine phosphoribosyltransferase (APRT) activity. The genetic basis of this phenotypic alteration is uncertain. It will be assumed that the APRT-phenotype can arise from point mutations in the gene for the enzyme as well as from stable epigenetic changes. This project is designed to distinguish between these two possibilities and identify nonsense suppressor genes and/or regulatory genes. A series of APRT- rodent cell clones will be selected and characterized. It is expected that some will owe their deficient phenotype to a nonsense mutation in the aprt gene or to an epigenetic mechanism involving the loss of function of regulatory genes. Concommitantly, a series of APRT- human cell clones will be selected, some of which may carry nonsense suppressor genes and/or genes capable of activating (derepressing) the rodent aprt gene. The human clones (suppressor clones) will be fused to the rodent clones (suppressor detector clones) in medium which selects for APRT plus cells. Rare combinations should yield a high frequency of APRT plus hybrids as a result of nonsense suppression or gene activation. These hybrids will be analyzed with regard to the species origin of the APRT expressed and whether or not it differs from the wild-type enzyme. This analysis will suggest the etiology of the APRT- phenotypes of the rodent and human parents. In particular, nonsense mutations in the aprt gene should be distinguished from epigenetic alterations. Finally, segregation analysis of the hybrid clones will permit the genetic mapping of human nonsense suppressor and/or regulatory genes.